Typical bar code decoding algorithms use hard thresholding to determine whether a given group of bars and spaces represents a particular character. These algorithms operate by first sensing the widths of bars and spaces in an input image and then normalizing these sensed widths in accordance with a nominal width "X" which is equivalent to the narrowest bar or space in the symbology being decoded. Next, each normalized width is thresholded and classified as being one of a plurality prototype widths. For example, in code 39 where each bar or space can either be narrow (1X) or wide (2-3X), each normalized width is thresholded against a value equivalent to 1.75X. If the normalized width is greater than 1.75X, then the element is determined to be wide, otherwise the element is determined to be narrow. A set of these thresholded widths representing a character in the symbol being decoded is then matched against a reference table which defines the character assignments for the symbology being decoded. This decoding algorithm for Code 39 is set forth in Appendix B of the Uniform Symbology Specification for Code 39 , dated 1986, published by the Automatic Identification Manufacturers, Inc., Pittsburgh, Pa.
Code 128 is also typically decoded using hard thresholding. In Code 128, the normalized width of each bar or space is either 1X, 2X, 3X, or 4X and each Code 128 character is comprised of three bars and three spaces. In Code 128, the normalized widths of the sensed bars and spaces are used to calculate the values T1, T2, T3, T4 which respectively correspond to the (i) distance from the leading edge of the first bar to the leading edge of the second bar, (ii) the distance from the trailing edge of the first bar to the trailing edge of the second bar, (iii) the distance from the leading edge of the second bar to the leading edge of the third bar, and (iv) the distance from the trailing edge of the second bar to the trailing edge of the third bar. Each calculated value T1, T2, T3, T4 is next normalized to the dimension X and then classified as falling within one of six potential width ranges. Finally, each value T1, T2, T3, T4 is determined to have a value equivalent to the midpoint of the range in which it has been classified. A set of these thresholded values T1, T2, T3, T4 representing a character in the symbol being decoded is then matched against a reference table which defines the character assignments for the symbology. This decoding algorithm for Code 128 is set forth in Appendix B of the Uniform Symbology Specification for Code 128, dated 1986, published by the Automatic Identification Manufacturers, Inc., Pittsburgh, Pa.
Hard thresholding techniques often produce errors when the normalized width value being thresholded falls near the edge of a thresholding range. This is particularly the case when low resolution images are used as the basis for calculating the normalized widths, because in such images the width of a bar or space in the bar code symbol being imaged may be equivalent in dimension to a single pixel in the image used to sense the symbol.
During the decoding of bar code symbols represented by low resolution images, it often happens that the bar code symbol cannot be successfully decoded by an initial scan through the symbol. Decoding systems recognize this condition when there has been a checksum error, or when a set of thresholded values representing one of the characters in the bar code symbol cannot be successfully matched against any of the entries in the decode reference table. In such situations, all the information representing the current scan of the bar code symbol is often thrown away, and the bar code symbol is evaluated by scanning it again (possibly through a different scan line) and attempting to evaluate the width measurements sensed from this further scan.
It is an object of the present invention to provide an improved system for decoding bar code symbols represented by low resolution images.
It is a further object of the present invention to provide a system for decoding bar code symbols that is more accurate and robust than hard thresholding methods.
It is a further object of the present invention to provide a system for combining information from multiple scan lines to enhance the decoding process in those cases where a bar code symbol cannot be successfully decoded by previous scan lines.
Further objects and advantages of the invention will become apparent from the description which follows.